Cameron Peake

Joining @financialstudio as a partner to invest in & support great fintech founders! @ryan_falvey techcrunch.com/2022/02/08/fi… #helloworldagain #fintech #fintechnews #VC

View from the firing squad! Thanks to @Shannon_Sibold for organizing the @AustinTech pitch panel alongside @MegBaldini Gavin Myers, and Anthony Marfisi. And to @TowItApp, Inevitable, @tomchepucavage, Manolo AI, @Mulgacloud, and @GlidTech

As agentic systems scale, the stakes become higher. Understanding agent identity will become one of the most important things we can do support this evolution. For anyone interested or curious, check out Beltic and our why we invested memo restive.com/blog/why-we-inve…

When AI was first taking off, there was a lot of hand-wringing (and falling stock prices) around the belief that the tech was a disruptive innovation for Google and the other tech giants. Three years in, these companies have embraced AI and it more accurately is a “sustaining innovation” according to Christensen's definition (and as evidenced by their recent growth rates). It seems to me that the industry more under threat is retail.  The latest example of this is Amazon blocking Muse from their platform this weekend. Muse threatens Amazon’s $68B+ ad business and its ability to influence cart size and checkout.  Retail in general has been cautious when it comes to agentic commerce. And the threat of disintermediation is real. However, Christensen's insight was never that incumbents fail to see disruption coming, it's that responding rationally to today's customers and today's profit pools is what sinks them. It feels inevitable that agents become more central in purchasing as we think about consumer preferences. The question is: Will this caution on the part of retail incumbents open the door to new entrants? Will those that optimize for agentic discovery and checkout pull ahead? Or will retailers slowly build in the right tools to take advantage of this consumer shift? Right now, it feels like the threat is coming faster than ever. Retailers have a chance to influence strategy, but that window is rapidly closing. gizmodo.com/amazon-brings-do…

There is no shortage of buzz around agentic payments, but the consumer war is what gets the attention. The bigger prize is in business payments. The volumes are bigger (over $100 trillion moves between businesses globally every year), no one likes their provider, there is massive fragmentation with little innovation, and businesses are actually willing to pay for efficiency. The proof is already out there. Ramp's AP agents flagged more than $1 million in fraudulent invoices in their first months, Mercury now issues cards directly to agents, and in our own portfolio, Natural is building the payments layer for B2B agents. A few takeaways for founders building here: - Don't just be the tollbooth. The work around each payment is a labor pool worth hundreds of billions a year. - Controls are a big part of the product. Humans set the bounds, agents operate inside them. - Given how early this space is, start with the early adopters, but look to find your path to where the money actually sits. - Ramp has been able to see volume because they own their environment. Building out agentic systems within an external org is starting from square 1. This will catch up, but if you own a network you have an advantage right now. In other words, the consumer battle will keep the headlines (for now), but the B2B version will mint the next payments platform. If you're building here, check out the deeper article and be in touch! restive.com/blog/the-real-pr…

Vanguard just made the largest acquisition in its 51-year history, paying a reported $4 billion for Altruist, the AI-native custodian serving nearly 5,000 independent advisory firms. Is it a good strategic move? Yes. But it's missing the point. The bull case is that this is Vanguard doing what Vanguard does. They spent five decades compressing costs toward zero, and they have become a complete behemoth. Over that same period the advisory fee barely moved. Bringing the low cost model to the last uncompressed fee in investing is a real opportunity (and a very Vanguard thing to do), and I can see them leveraging Altruist to lower advisor fees, improve efficiency, and credibly expand Vanguard's product suite. This is a bet that relationships will still matter, and an advisor is at the core of financial planning. The caveat is that the bet ignores the state of tech today. Will there be human advisors in twenty years? Will they serve everyone, or only the top 5% of households who want white-glove service? The much more Vanguard play would have been building a state-of-the-art AI advisor that needs no humans at all, the way they once built index funds rather than buying active managers. But the other hallmark of Vanguard, right alongside low costs, has been a consistent underinvestment in technology, and you cannot build the future of advice on a stack that has struggled to keep a mobile app running. With that option not really on the table, betting on the future of advisory relationships was their rational strategic move. Vanguard may execute it exceptionally well. It just may be the wrong bet. PS- this is a fun flashback to our panel at Go Restive with @ebloch and @Eden__Ovadia earlier this year! News: lnkd.in/g7VJpbTg

I think the most interesting angle on Stripe's OpenRouter acquisition is not where the company is today but what it could be in the future. Is orchestration tech a commodity? Yes. Is the data and network they bought valuable? Yes. But critically, OpenRouter is arguably the largest aggregator of the long tail of developers working on agentic systems. And as we think about agents moving more directly into commerce, it's an important top-of-funnel source to position Stripe as the default for agentic payments (alongside their cozy partnerships with OpenAI and Anthropic). The defensive future is just as important. OpenRouter aggregates demand behind a funded account, meters every token consumed, and settles with providers on the other side, and that combination is a payments network in embryo. Had OpenRouter reached its full potential as an independent company, it could have become a significant threat. Bridge followed the same logic when Stripe bought it: a capability purchase that also removed a rail that could have routed around them. Is $7B a huge price tag for a relatively new company? Absolutely. But Stripe is betting on the future, and would rather OpenRouter realize that potential inside Stripe than against it. News: techcrunch.com/2026/08/16/st…

Many merchants are still keeping AI agents out of the shopping experience. This morning, @shopify reported what happens when you let them in. Shopify's agentic infrastructure has been some of the smartest I've seen at scale. Their Catalog structures over a billion products that agents can read and transact against and they co-developed Google's Universal Commerce Protocol to standardize how agents buy across the web. Their Q2 numbers, drawn from millions of storefronts, supply the data on where this market is going. - AI-driven traffic and AI-generated orders each tripled year over year - Orders from AI channels were 2x as likely to come from a brand-new customer as orders from organic search - AI search built on structured catalog data converted 2x as well as AI search built on scraped data - 75% of AI-attributed purchases came from outside the top 100 product categories What do those numbers say together? AI is driving growth and discovery. It introduces shoppers to merchants they'd never have found, and pulls products out of the long tail that search never surfaced. And merchants who invest in making their inventory discoverable convert that traffic at far better rates than those who haven't. In other words, the merchants who win the next era of discovery will not be the ones with the best chatbots. They will be the ones machines can see. Shopify's earnings: retailtouchpoints.com/news/s… Our take on the machine layer: restive.com/blog/the-machine…

Anyone building in AI financial services should check this out! It outlines what we got right and wrong from our predictions a year ago, and what to look for in the year ahead

now • Michaels launched an AI shopping assistant this week. It has handled 75,000 conversations since May. Michaels.com gets more visits than that before lunch. Retailers have an opportunity to do so much more with what AI and agentic commerce present, and to advance their business. But for so many, AI feels like a performative exercise for shareholders as opposed to a paradigm shift. In fact many large retailers are actively blocking AI agents from the shopping experience, even as shoppers move their discovery into those same tools. In a December survey of 1,030 consumers, 55% said they use AI for product research at least weekly, and half had already made a purchase after researching with AI. A few retailers seem to be approaching AI in novel ways: @Target and @Walmart have been embracing partnerships with ChatGPT to support inventory and discoverability on the platform. @Shopify has been building the structured catalog feeds that agents actually read. But even these are incremental steps. Full agentic shopping still needs infrastructure that mostly does not exist yet, including machine-readable catalogs, agent identity, and payment rails built for software that buys. We wrote about that missing layer here: lnkd.in/gCiRR5jA I will likely brainstorm my kid's next birthday party in Claude. But I am not holding my breath that I could direct an agent to actually buy the decorations at Michaels. In other words, the winners in agentic commerce will not be the retailers with the best chatbots. They will be the ones who reconceptualize what it means to partner, discover, and buy. lnkd.in/g7Zu4uMb

I spent my early career launching businesses in emerging markets. Lately I've been getting déjà vu watching hardtech. Take CoreWeave. It buys GPUs and borrows billions against its customer contracts, with debt rated off the customer's credit, not its own. I watched pay-as-you-go solar companies build the exact same machine a decade ago. In India and Kenya, d.light sells solar systems to off-grid households in installments, then securitizes the receivables. The hardware is collateral, the contract is the asset, and the "hardware company" is really a lending business with a machine out front. Once you've seen it in Nairobi, you see it everywhere in hardtech: In Taiwan, Gogoro stopped selling scooter batteries and rents them by the swap. That annuity now generates ~$150M/yr in growing recurring revenue. In the US, Formic doesn't sell factory robots. It charges by the hour, only when the robot hits agreed output. In Texas, Base Power doesn't just sell home batteries. It makes households co-owners of a power plant and shares the grid revenue. Just like M-KOPA and Okra Solar did with village solar. Here's why this matters to us at a fintech fund. The markets where I worked were brutal: customers earning $2/day, capital costing 15-25%. The hardware businesses mostly broke even. The financial layer scaled anyway. The model survived its stress test in the hardest markets on earth. Now it runs with cheap capital in easier markets, and that's a rerun worth watching. I wrote up the full pattern: restive.com/blog/everything-…

Riding on the heels of @getpiehq fundraising announcement yesterday, we wanted to share why we invested in Syed, @akhilmantrip, and the vision they laid out to help small businesses grow. The basic observation is rooted in what Syed and Akhil saw while working at Toast and Square (and what I saw in my last startup): while there are many nice-to-have backoffice or efficiency products in the market, ultimately SMBs live, die, or thrive by the growth they can drive to their businesses. We've since witnessed Syed's unrivaled skill selling to SMBs and Akhil's brilliance on product. Perhaps most important, we've seen Pie become the trusted partner to help SMBs use AI for growth - whether that is voice, LLMs, or GEO. Big congrats to the team on this milestone! And if you are an SMB looking to grow, check out Pie. lnkd.in/gyyeAZAG

Coinbase just launched an SEC-registered AI financial advisor. It's an exciting headline, but structurally it's not all that novel- robo-advisers have been registered investment advisers for fifteen years. That said, I do think it's a big deal. Not because of the headlines, but because of what it represents for the industry. We know the capability for AI to do the analytical core of wealth management is there. What this represents is democratization. Today, personalized financial advice is gated by account size. If your portfolio is too small, no human's time is worth it, so you get nothing. AI has no minimum. It can serve the person with $10,000 as carefully as the person with $5M. For tens of millions of people, that's the difference between no advice and real advice. This was the core of what @ebloch was building at Hiro. The challenge to the industry isn't "how do we keep AI out until the rules catch up." It's "how do we use the existing framework to bring AI in". As Coinbase just showed, this is very doable. The regulatory moat for advice is thinner than people pretend. And every firm hiding behind it is making two bets at once: that millions of people will keep going without advice they could finally access, and that no one else will walk through the door Coinbase just opened. Both are bad bets